Common shareholders buying stock in Anthropic have no say over the company’s board of directors.
The firm was founded in 2021 with the aim of ensuring humanity controls AI and benefits from it over the long term. As the company prepares for its initial public offering, that mission creates a governance model that excludes public investors from most major decisions.
Complex structures allowing founders to control public companies without holding the majority of shares are not new in Silicon Valley. Anthropic’s focus on AI safety has produced a corporate governance system that leaves ordinary investors with little influence.
Anthropic is a public benefit company. This legal status allows the corporation to prioritise a public mission alongside profit. The arrangement protects the board from being sued by angry investors if it makes decisions about AI safety that are not driven solely by shareholder value.
The current board is chaired by Anthropic president Daniela Amodei. It includes CEO Dario Amodei, her brother, Netflix cofounder Reed Hastings, former Microsoft and GM CFO Christopher Liddell, Novartis CEO Vas Narasimhan, and Spark Capital partner Yasmin Razavi. There is one vacancy on the board.
A Long-Term Benefit Trust holds a separate class of stock that gives it the power to elect and remove a majority of the board, known as Class T directors. The trust is chaired by Neil Shah, a global health expert and former managing director at GiveWell. Its members include former Federal Reserve chair Ben Bernanke, national security expert Richard Fontaine, and Harvard physicist Christopher Stubbs. There is one vacancy on the trust, which existing members will fill.
The seven founders jointly own a special class of stock. This grants them 50.1% of shareholder voting power, allowing them to elect three members of the board of directors. If there is no majority among the founders, their shares are voted in proportion with holders of common stock.
The board can also check the Long-Term Benefit Trust. When the trust nominates a “Specified” director or removes one without cause, the board’s nominating committee can decline to approve the move. This throws the decision to the company’s employees for a vote.
Here is how the structure functions in practice.
Predicting the outcome of such a complex system is difficult. A similar structure was recently challenged at OpenAI. The board briefly fired CEO Sam Altman over concerns regarding his candour. Employee loyalty to Altman caused the board to back down and reinstate him. A new governance structure was then put in place. Anthropic’s model seems to formalise that kind of employee power.
Cracks in the founders’ unity could allow common shareholders to gain more influence than expected. Employees can overrule the Long-Term Benefit Trust on board seats if they back the board over the assembled experts.
The founders’ control is temporary. Ten years after the IPO, they can choose to end their control of the company. Alternatively, control ends automatically when fewer than three founders remain. At that point, the Long-Term Benefit Trust continues until 85% of common shareholders vote to end it, or the trust decides to step down with board approval.
If and when neither special class of shareholder exists any more, common shareholders finally get control.
What it means
For people making things with the technology, the practical effect is that the board retains authority over safety and strategic direction regardless of market performance. The founders keep significant leverage for a decade, but the system includes a built-in mechanism for employees to intervene if the board and the Long-Term Benefit Trust clash. Public investors remain passive observers of these decisions until the founders fade or the trust is dissolved.
